Alzheimer's disease afflicts over 5 million Americans and is the 6th leading cause of death in the U.S. To date, advanced dementia research has largely focused on describing the end-of-life experience of patients with this disease. Designing and testing interventions targeting those opportunities is the current research priority for this field. Advance care planning (ACP) is the most consistent modifiable factor associated with better palliative care outcomes in advanced dementia. The opportunity for ACP is exceptional in advanced dementia but often inadequate. Thus, advanced dementia patients often get aggressive interventions that may be inconsistent with preferences and of little clinical benefit. Recent work has particularly underscored the need to avoid unwanted and unnecessary hospitalizations among these patients. Traditional ACP primarily relies on ad hoc verbal descriptions of hypothetical health states and treatments. This approach is limited because complex scenarios are difficult to envision, information from providers is inconsistent, and verbal explanations are hindered by literacy and language barriers. To address these shortcomings, the co-PIs have developed video decision support tools for ACP and shown their efficacy in several randomized controlled trials (RCTs) in out- patient settings. The over-riding goal of the EVINCE (Educational Video to Improve Nursing home Care in End- stage dementia) study is to conduct a cluster RCT of an ACP intervention vs. control among 360 nursing home (NH) residents with advanced dementia (N=180/arm) in 20 matched NHs (10 intervention/10 control). At baseline, proxies in the intervention NHs will view a video ACP decision support tool. After viewing the video, their preferred level of care for the resident (comfort care, limited care, life prolonging care, or uncertain) will be ascertained and communicated to the primary care team. Proxies in the control NHs will experience usual ACP practices. The Aims are: Aim 1: To compare proxies' preferences for the residents' level of care in the intervention vs. control NHs at baseline (10-minutes post video in intervention NHs), 3, 6, 9 and 12 months. Aim 2: To compare ACP among residents in the intervention vs. control NHs at 3, 6, 9, and 12 months as measured by documented: 1. Decisions to forego hospitalization; 2. Decisions to forego other treatments (tube- feeding, parenteral therapy), and 3. Goal of care discussions; and Aim 3: To compare hospital transfers and other burdensome treatments (tube-feeding, parenteral therapy) over 12 months among residents in intervention vs. control NHs. A decision to forego hospital transfers by 6 months will be the primary outcome of this RCT. Impact: Better ACP is a key opportunity to improve advance dementia care. Video decision support is a practical, evidence-based, and innovative approach to ACP. If this RCT is successful, this will be one of the first rigorously tested interventions shown to improve outcomes in advanced dementia. This work could have significant clinical and policy implications for the millions of Americans dying with ths disease by promoting care that is more consistent with their preferences and that is less burdensome and costly.